Scene from inside: The representation of Observer's space in high-level visual cortex

NEUROPSYCHOLOGIA

卷:161

Article

Chaisilprungraung, Thitaporn¹  Park, Soojin² 

[1] Johns Hopkins Univ, Dept Cognit Sci, Baltimore, MD 21218 USA

[2] Yonsei Univ, Dept Psychol, Seoul, South Korea

关键词: Scene;    Landscape;    Spatial boundary;    PPA;    Mid-level visual feature;    MVPA;   

DOI  :  10.1016/j.neuropsychologia.2021.108010

来源: Elsevier

【 摘 要 】

Human observers are remarkably adept at perceiving and interacting with visual stimuli around them. Compared to visual stimuli like objects or faces, scenes are unique in that they provide enclosures for observers. An observer looks at a scene by being physically inside the scene. The current research explored this unique observer-scene relationship by studying the neural representation of scenes' spatial boundaries. Previous studies hypothesized that scenes' boundaries were processed in sets of high-level visual cortices. Notably, the parahippocampal place area (PPA), exhibited neural sensitivity to scenes that had closed vs. open spatial boundaries (Kravitz et al., 2011; Park et al., 2011). We asked whether this sensitivity reflected the openness of landscape (e.g., forest vs. beach), or the openness of the environment immediately surrounding the observer (i.e., whether a scene was viewed from inside vs. outside a room). Across two human fMRI experiments, we found that the PPA, as well as another well-known navigation-processing area, the occipital place area (OPA), processed scenes' boundaries according to the observer's space rather than the landscape. Moreover, we found that the PPA's activation pattern was susceptible to manipulations involving mid-level perceptual properties of scenes (e.g., rectilinear pattern of window frames), while the OPA's response was not. Our results have important implications for research in visual scene processing and suggest an important role of an observer's location in representing the spatial boundary, beyond the low-level visual input of a landscape.

【 授权许可】

Free